i think AMMs will be fragmented based on what type of pairs their design optimizes for, so it becomes an LP decision, not a user decision
users will just use aggregators
imo larger users really dont care about gas optimization, tx fees are just the cost of doing business
Conversation
Replying to
One thing that I think is cool -- you can aggregate all AMMs naively together by having them all achieve their curves by providing in the same orderbook.
3
22
This Tweet was deleted by the Tweet author. Learn more
so imagine that instead of going directly to an AMM:
1) AMM wants x*y=k; so the AMM places offers out to mimic that curve; same with bids
2) when an offer gets filled, it replaces with a bid, and vice versa
3) to trade on the AMM you just trade on the orderbook
1
1
12
Sounds cool. But eth network clogging can be fatal to this design? Could sth like this be built at l3 level maybe?
1
1
I never think the fee is the fee the most important factor(not saying it is not important). The speed of transaction is probably more crucial in this case?
1
3
How many full nodes you have here, sir? Independent ones
1
Interesting. Is it some variant of bft/dbft? I heard performance usually drops dramatically >20 nodes.
1
You can get exponential growth in the number of nodes that have validated (each one spreading to 200 more, etc.), so the scaling isn't too bad.
Plus proof of history/time can help resolve the ordering of transactions.
1
Show replies